A novel quantum genetic algorithm for detection sonar image

According to the manifold feature of sonar image data, a novel quantum genetic algorithm for detection sonar image is proposed in this paper. K-means block clustering algorithm is used to reduce the space of sonar image data. On this basis, in order to reflect the global consistency of the data, manifold distance is developed as the new similarity measurement of fitness function in quantum genetic algorithm. In the process of specific detection, K-means block clustering algorithm as local search method is added to the evolution of quantum genetic algorithm, the combination of global search and local search finish evolution to avoid early maturity of quantum genetic algorithm. Experiment results of the artificial data sets and real sonar images demonstrate that the proposed algorithm has certain accuracy and adaptability.